Background:
Hospital readmission within 30 days of discharge is an area of emphasis of health care reform. Recent reports reveal that readmissions are costly and potentially preventable in many cases. Limited research is available on what factors place a patient a risk and if targeted interventions can reduce the observed readmission rate. Our objectives were to determine if there are patient factors that correlate to 30‐day readmission rates, and if a cumulative set of factors correlates to increased observed rates.
Methods:
We performed a retrospective, observational cohort study. Patients were eligible if they were discharged from general medicine services at a single academic medical center between April 1, 2010, and June 30, 2010. Exclusion criteria were death during hospitalization, leaving AMA, and encounters within 30 days of index hospitalization. As a participant in Project BOOST, much emphasis was placed on the target screening tool criteria. Patient records were evaluated for: age > 65; problem diagnosis based on coded diagnoses of acute or chronic CHF, acute or chronic obstructive pulmonary disease, acute pneumonia, acute stroke, active cancer, and acute or chronic diabetes mellitus; problem medications (anticoagulants, insulin, aspirin and clopidogrel, digoxin, or scheduled narcotics); polypharmacy (defined as ≥ 10 medications on the discharge orders); prior hospitalization at our institution within 6 months; and coded diagnosis of depression. We identified readmissions from administrative data and confirmed by chart review. Charts were scored for 6 primary risk factors, and an overall score was calculated. Chi‐square analysis was performed. We also assessed the value of cutoff scores of 0, 1‐3, and 4‐6.
Results:
Eight hundred and fifty‐eight cases met study criteria. The observed 30‐day readmission rate was 15.96%. Table 1 shows the prevalence of the 6 factors studied, including observed 30‐day readmission rate, calculated odds ratios, and P values. Table 2 shows the correlation of a range of scores and observed read‐mission rates. Problem diagnosis, problem medication, polypharmacy, and prior admission within 6 months were significant individual risk factors correlating with observed 30‐day readmissions in our study. A score of 1‐3 and >4 factors were indicative of a significantly higher risk for 30‐day readmission than a score of 0.
Table 2.
Conclusions:
Select patient factors and a screening score obtainable near the time of discharge can be used to identify patients at higher risk for 30day readmission. These findings may be useful to determine targeted use of interventions to reduce readmission risk.
Disclosures:
A. Pal ‐ none; M. Taha ‐ none; A. Merando ‐ none; M. McDonald ‐ none; C. Wesselman ‐ none